1.6 News & Views MH NEW

583 syndrome: behavioural changes, decreased muscle strength, heart defects, annular pancreas, aganglionic megacolon, immune defects and placental vascular abnormalities. So Arron et al. looked for genes in the DSCR that could regulate NFATc. They identified DYRK1A, which encodes a kinase, and DSCR1, which encodes a known inhibitor of calcineurin. The DYRK1A and DSCR1 proteins acted synergistically to block NFATc-regulated gene expression. The authors found that DYRK1A acts as a priming kinase that enables additional phosphorylation of NFATc by another kinase, glycogen synthase kinase 3, leading to NFATc inactivation. Furthermore, mice genetically engineered to have increased levels of DYRK1A alone, or of both DYRK1A and DSCR1, had abnormal heart-valve development. As expected, in these mice NFATc was mostly phosphorylated and found in the cytoplasm (Fig. 1b). Gwack et al. reached a similar conclusion on the role of DYRK1A in regulating NFAT signalling by using an innovative approach in fruitfly cells to identify regulators of the mammalian NFAT pathway. Although these cells do not themselves possess NFATc factors, the pathway regulating their movement is present. This provided the basis for a genomewide screen using interfering RNAs to inhibit the expression of specific genes. The authors identified DYRK1A and its relative DYRK2 as direct regulators of NFATc phosphorylation, and, like Arron et al., they showed that DYRK1A has a priming function. Gwack et al. note that DYRK1A and DSCR1 occur in the DSCR, and suggest that their findings might aid the understanding of the immunological and neurological defects in Down’s syndrome. But does perturbation of the NFAT system explain the many developmental abnormalities of Down’s syndrome? The difficulty in establishing this, as Arron et al. point out, is that the various genetically engineered mice show different developmental defects. The trisomic Ts65Dn and Ts1Cje mice, which express extra copies of DYRK1A and DSCR1 (ref. 5), show few of the abnormalities found in mice deficient in the various NFATc factors. But the trisomic mice also show few of the abnormalities of Down’s syndrome, although they do have cellular abnormalities that may be highly relevant in studying the condition. Such discrepancies may result from subtle variations in the expression of components of the NFATc regulatory system that occur because of the ways the relevant genes have been introduced or knocked out. Similarly, the discrepancies between Down’s syndrome in humans and the mouse models may derive from species differences in gene expression and regulation. Whatever their causes, these discrepancies make it more difficult to analyse directly the components of Down’s syndrome that are not replicated in the trisomic mouse models. However, one feature that occurs in the human syndrome, NFATc-deficient mice and trisomic mice is the abnormalities of the skull and jawbone that provided the original impetus for the search by Arron et al.. So a straightforward way to test whether increased DYRK1A and/or DSCR1 expression is the cause of these defects would be to reduce the number of copies of either or both genes from three to two by mating trisomic mice with mice lacking the genes. A similar approach has already been used to explore the role of the APP protein in the brains of trisomic mice. Increased DYRK1A activity was already suspected to be a factor underlying the cognitive and muscular deficits in Down’s syndrome. The work of Arron et al. and Gwack et al. suggests another way that increased activity of DYRK1A and DSCR1 may contribute not only to mental retardation, but also to many other features of Down’s syndrome. ■ Charles J. Epstein is in the Department of Pediatrics and the Center for Human Genetics, University of California, San Francisco, Rock Hall 584B, 1550 4th Street, San Francisco, California 94143-2911, USA. e-mail: [email protected]